Importing Important Libraries¶
In [1]:
import numpy as np
import pandas as pd
import missingno as mn
import plotly.express as px
import plotly.graph_objects as go
from wordcloud import WordCloud, ImageColorGenerator
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings("ignore")
Importing data and creating a deep copy¶
In [2]:
videoGameData = pd.read_csv("vgsales.csv",index_col=0)
In [3]:
data = videoGameData.copy()
Examining the Data¶
In [4]:
data.head()
Out[4]:
In [5]:
data.shape
Out[5]:
In [6]:
data.size
Out[6]:
In [7]:
data.describe()
Out[7]:
In [8]:
data.info()
Checking Duplicated values¶
In [9]:
data.duplicated().sum()
Out[9]:
In [10]:
data.drop_duplicates(keep = 'first',inplace = True)
Checking Null value¶
In [11]:
data.isnull().mean() * 100
Out[11]:
In [12]:
mn.matrix(data)
Out[12]:
We dont have much null values, so i would prefer dropping over synthesizing¶
In [13]:
data.dropna(inplace=True)
Year Datatype should be integer¶
In [14]:
data['Year']=data['Year'].astype('int64')
Exploratory Data Analysis¶
1.Top Grossing Games¶
In [15]:
GamesRev = data.groupby('Name')['Global_Sales','NA_Sales','EU_Sales','JP_Sales','Other_Sales'].sum().sort_values('Global_Sales',ascending = False)[:10]
GamesRev.drop('Global_Sales',axis=1,inplace=True)
fig = px.bar(GamesRev,
labels={
"variable": "Region",
"value": "Sales(in Million)",
"Name": "Game"},
title = 'Top Grossing Game in different Region')
fig.update_layout(
title={'y':0.9,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.update
fig.show()
Wii Sports was played the most with highest global sales followed by evergreen GTA V & Super Mario Bros¶
2.Region Wise Sales Distribution¶
In [16]:
region =['NA_Sales','EU_Sales','JP_Sales','Other_Sales']
sales = [data[i].sum() for i in region]
fig = px.pie(names=region,values=sales)
fig.update_traces(rotation=90, pull=[0.06,0.06,0.06,0.06], textinfo="percent+label")
fig.show()
North America constitutes 49.1% of the global game sales¶
3.Top Grossing Publisher¶
In [17]:
publisher = data['Publisher'].value_counts()
publisher2 = data.groupby('Publisher')['Global_Sales'].sum().sort_values(ascending=False)[:10].reset_index()
publisher = publisher.reindex(index=publisher2['Publisher'])
publisher2['Games Published'] = publisher.values[:10]
fig = px.bar(publisher2 , x =publisher2['Publisher'],y=publisher2['Global_Sales'],color=publisher2['Games Published'],
labels={"Global_Sales": "Global Sales"},
title = 'Top Grossing Publisher with number of Publications')
fig.update_layout(
title={'y':0.9,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
4. Top Publishers Top 10 Grossing Games¶
In [18]:
topPublisher = publisher2['Publisher'][:5]
publisher = data.groupby(['Publisher'])
fig = go.Figure()
for pub in topPublisher:
pubget = publisher.get_group(pub).groupby('Name')['Global_Sales'].sum().reset_index().sort_values('Global_Sales',ascending = False)[:10]
fig.add_trace(
go.Bar(x=pubget['Name'],
y=pubget['Global_Sales'],
name=pub,
visible= True if pub == 'Nintendo' else False,
marker={'color': pubget['Global_Sales'],'colorscale': 'Portland'}))
fig.update_layout(xaxis_title="Games",
yaxis_title="Global Sales")
fig.update_layout(
updatemenus=[
dict(
type="buttons",
direction="right",
active=0,
x=0.85,
y=1.1,
buttons=list([
dict(label=topPublisher[0],
method="update",
args=[{"visible": [True, False,False, False, False]},
{"title": "Top 10 Games per Publisher"}]),
dict(label=topPublisher[1],
method="update",
args=[{"visible": [False,True, False, False, False]},
{"title": "Top 10 Games per Publisher"}]),
dict(label=topPublisher[2],
method="update",
args=[{"visible": [False,False, True, False, False]},
{"title": "Top 10 Games per Publisher"}]),
dict(label=topPublisher[3],
method="update",
args=[{"visible": [False,False, False, True, False]},
{"title": "Top 10 Games per Publisher"}]),
dict(label=topPublisher[4],
method="update",
args=[{"visible": [False,False, False, False, True]},
{"title": "Top 10 Games per Publisher"}]),
]),
)
])
# Set title
fig.update_layout(
title_text="Top 10 Games per Publisher",
xaxis_domain=[0.05, 1.0],
title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28),
annotations=[
dict(text="Publisher :", showarrow=False,
x=0.2, y=1.1, yref="paper", align="left")
]
)
fig.show()
5.Most Popular Genre¶
In [19]:
text = list(set(data['Genre']))
plt.rcParams['figure.figsize'] = (15,15)
wordcloud = WordCloud(max_font_size=50, max_words=100,background_color="black").generate(str(text))
plt.imshow(wordcloud,interpolation="bilinear")
plt.axis("off")
plt.show()
In [20]:
genre = data['Genre'].value_counts()
fig = px.pie(genre, values = genre.values , names=genre.index,title= "Genre Distribution")
fig.update_traces(textinfo="percent+label")
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
Most Games published were of the ACTION genre followed by SPORTS and MISC¶
6.Top Grossing Genre¶
In [21]:
genreSales = data.groupby('Genre')['Global_Sales'].sum()
fig = px.bar(genreSales,orientation='h',labels={"value": "Global Sales","variable":"region"},title = 'Top Grossing Genre')
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
ACTION Games are the highest grossers of all time followed by SPORTS and SHOOTER Games¶
7.Most Favourite Platform¶
In [22]:
text = list(set(data['Platform']))
plt.rcParams['figure.figsize'] = (15,15)
wordcloud = WordCloud(max_font_size=50, max_words=100,background_color="black").generate(str(text))
plt.imshow(wordcloud,interpolation="bilinear")
plt.axis("off")
plt.show()
In [23]:
platform = data['Platform'].value_counts()[:10]
fig = px.pie(platform,names=platform.index,values=platform.values,hole=.3,title='Platform Distribution')
fig.update_traces(textinfo="percent+label",pull=[0.1,0.2,0.15,0.06,0.06,0.06,0.06,0.06,0.06,0.06])
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
Mostly Games were published for DS , PS2 and PS3¶
8.Top Grossing Platform¶
In [24]:
platform = data.groupby(['Year','Platform'])['Global_Sales'].sum().reset_index()
platform = platform[(platform['Year']>=2006) & (platform['Year']<=2015)]
platform = platform.loc[platform['Platform'].isin(['DS', 'PS2', 'PS3', 'Wii', 'X360'])]
fig = px.bar(platform,x='Platform',y='Global_Sales',labels={"Global_Sales": "Global Sales"},
title = 'Top Grossing Platform over 10 Years',animation_frame='Year',
animation_group='Platform', color='Platform')
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
9.Sales Distribution over Years¶
In [25]:
yearSales = data.groupby('Year')['Global_Sales','NA_Sales','EU_Sales','JP_Sales','Other_Sales'].sum()
fig = px.line(yearSales,title='Region-wise Sales Distribution over Years',
labels={"value": "Sales(in Million)","variable":"Region"} )
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
Video Game Sales witnessed Global High between 2006 to 2011¶
10.Sales relation with number of Game Published¶
In [26]:
year = data['Year'].value_counts()
year2 = data.groupby('Year')['Global_Sales'].sum().sort_values(ascending=False).reset_index()
year = year.reindex(index=year2['Year'])
year2['Games Published'] = year.values
fig = px.bar(year2 , x =year2['Year'],y=year2['Global_Sales'],color=year2['Games Published'],
title='Sales Relation with Number of Game Published',labels={"Global_Sales":"Global Sales"})
fig.update_layout(title={'y':0.95,
'x':0.5,
'xanchor': 'center',
'yanchor': 'top'},
titlefont=dict(size =28))
fig.show()
Video Game Industry witnessed a boom between 2006-2011 ,when more than 1000 games were published every year with Global Sales hitting all time high.¶
In [ ]: